The World CO2 Emission

What should we do?

Greenhouse gases are gases in the earth's atmosphere that trap heat. They emitted by human activities strengthen the greenhouse effect, contributing to climate change (U.S.Environmental Protection Agency). Carbon dioxide (CO2) is the primary greenhouse gas emitted through human activities. Our group applied several exploratory data analysis and visualization techniques on data sets related to GDP, global tonnes of CO2 emissions (tCO2) / Greenhouse Gas emissions (GHG), forest coverage, and energy use, both globally and by country. The overall goal of this project is to develop new insights about the relationships between these topics, as well as to better understand their behavior over time.

Overview

Global energy-related CO2 emissions grew by 0.9% or 321 Mt in 2022, reaching a new high of over 36.8 Gt (IEA). Since 2000, the sharp acceleration in CO2 emissions to more than a 3% increase per year have been due to accelerated economic growth. Table 1, found below, shows a continent-level summary of CO2 emissions. We can see that Asia’s CO2 emissions in 2019 are almost triple its values in 1990, whereas there is actually a negative net change in Europe between the same two years (Table 1). Our initial thought was that countries might have much more varied emission behavior than what we see on a continental-level, and this is part of the reason why we decided to look at the World CO2 emissions dataset. This dataset contains CO2 emissions of 195 countries from 1990 to 2020. It is a well-known dataset that has already been explored by many scientists and studies. We have tried to use different types of visualizations, including interactive and linked plots, while also incorporating other datasets to develop new insights. The idea is to use visualization to study the relationships between CO2 emissions and country level data related to it, such as forest cover, GDP, sectors, and energy use. Ultimately, we hope that we can provide insights about historical CO2 emissions and how countries take different actions to reduce or slow the increasing of CO2 emissions.

Table 1. CO2 emission summary by continent. The difference in net change from 1990 to 2019 among the different continents implies variation among CO2 emissions in different countries.
CO2 Emission by Continent (tonnes)
From 1990 to 2019
Continent 1990 Emissions 2019 Emissions Net Change
Africa 2.19B 3.78B 1.59B
Asia 9.85B 26.61B 16.76B
Europe 8.76B 6.24B −2.53B
North America 6.82B 7.71B 890.25M
Oceania 638.99M 817.18M 178.19M
South America 2.98B 3.11B 124.21M
Source: Our World in Data


How CO2 emission differ by country and other factors?

Figure 1. Linked view of overview of the dataset, choropleth map of CO2 emission and country level time series data including forest cover, GDP per capita and energy use per capita. Three leading countries, U.S.,China, and Russia, contribute over 50% of the world CO2 emission. Each country has various trend of forest cover, GDP per capita and energy use per capita over time (1990-2020). In this two-level linked plots, by clicking the countries, the bar plots are filtered to the specific country. The users can also chose the year of interest in any bar plots. Consequently, the same year of other plots would highlighted accordingly.
35
Gigaton
3
Leading countries
~50%
Emission

Figure 1 is an overview of our merged dataset. It contains CO2 emissions, forest percentage cover, GDP per capita, and energy use per capita annual time series data from 1990 to 2020. We are using linked/interactive choropleth map to help users discover the dataset. From the choropleth map, we can see there are three leading countries, China, the United States, and Russia, whose CO2 emissions took ~50% of the World's total emissions in 2020. Each country has unique trends of other factors. For example, China has an increasing trend of forest coverage. This led to our further analysis of the details of each factor. Another interesting thing we found from this figure is that the CO2 emissions stacked bar plot by country show that U.S. and China switched their roles around 2006. Before 2006, the U.S. had higher emissions, while after 2006, China took place.

This view is generated using Altair package in python. The color is in log scale so that the large range of the CO2 emissions can be presented well on the map. The color scheme is chosen based on our green theme. This figure can also be considered an innovative view due to it being a two-level linked plot. By clicking the countries, the bar plots are filtered to the specific country. The users can also choose the year of interest in any bar plot. Consequently, the same year of other plots would be highlighted accordingly.

Emission by sectors

  • What type of countries have higher CO2 emissions? Where do they come from?

    From Figure 1, we can see the CO2 emissions varied by countries. Here, we further examine if there are patterns of country type and CO2 emissions.

    This plot was made in R using Plotly package. There is one view that displays net change between the start and end years of the data, and to obtain the change we subtracted the most recent year’s values from the first year’s values. This was repeated for each economic sector and each income group to see if one sector/group was most responsible for emission trends. The color of each income group was selected to match the coloring of other plots that use the same variable (i.e. GDP in Figure 1 and Figure 4).

    Figure 2. CO2 emission by sectors and country income. Upper-Middle income countries contribute to CO2 emissions the most with high usage of electricity and heat. Low-income countries with inadequate technology emit CO2 mainly from agriculture.

Emission VS Forest

  • Figure 3. Joint behavior of emissions and forest cover of the top five countries with the most forest cover. Brazil is the only country in these five countries to show a negative relationship between forest cover change and CO2 emissions, which indicates, even theoretically increasing forest coverage can help to absorb CO2, there are other major factors affect CO2 emissions making the forest coverage change impacts on CO2 emissions negligible.


    Does forest strategy helps?

    From Figure 2, we can see the CO2 emissions are counted as negative values for forests when calculating carbon footprint. This is because forests sequester (or absorb) and store carbon dioxide from the atmosphere, helping reduce greenhouse gas emissions (source). In addition, deforestation is one of the main human activities that account for climate change and CO2 emissions. Thus, we further examine if increasing forest coverage can certainly help to reduce CO2 emissions in Figure 3.

    This plot was made in R using the Plotly graphing library. Each dropdown menu option is an individual trace that was added to the plotting object, and the menu toggles the visibility of the plots based on user input. The left axis corresponds to emissions, the solid line, while the right axis corresponds to forest cover, the dashed line. For each country, the estimated amount of trees that correspond to a 1% change in forest cover is provided as context. The aesthetic elements were changed to minimize unnecessary visual information.

Emission and GDP

How emission change with GDP?
Figure 4. Linked view of historical CO2 emissions change with GDP. Before the year of 2010, high-income countries have always been the main contributors to CO2 emissions. This trend changed after 2010, with upper-middle-income countries surpassing high-income countries to become the largest contributor to global CO2 emissions in terms of total emissions.

Figure 1 shows positive relationship between GDP and CO2 emissions. In addition, Figure 2 shows country income type is also affect the CO2 emissions distributed by sector. Here, we conduct Figure 4 to see the detailed change of CO2 emissions with GDP.

Figure 4 is a linked scatter plot with a horizontal bar plot at the bottom developed by Plotly package in python. The plot contains a slider with animation. Data from 1990 to 2019 will be displayed after the click. The scatter plot takes the emission of CO2 in a country on the x-axis, the emission of CO2 per capita in a country on the y-axis. The bubble size represents the population of a specific country, and the color of the bubbles represents the income group corresponding to a country. In the bar plot, the x-axis represents the overall CO2 emission number, and the y-axis represents different income groups. The plot is interactive, and hovering over the bubbles displays detailed information about each data point, such as country, total emission, emission per capita, income group, and population. And hovering over the bar chart displays detailed information about the specific income group and the overall emission of this group.

The scatter plot was inspired by the “gapminder” plot but with improvements by adding a horizontal bar chart to carry more information. The color encodes to correspond with income groups. Bright red was used to represent a high income group, while light green is used to represent a lower-income group, which matches Figure 2. The initial development of the plot used bokeh package, but due to the requirement of a server for the interactive bokeh plots, it was ultimately switched to the plotly package.

Before the year of 2010, high income countries have always been the main contributors to CO2 emissions. As shown in the plot, the small-sized bubbles are always located above others, which indicates that the high-income countries have smaller populations but higher emissions per capita. Lower-middle income countries, with their large population size, have consistently ranked second in total CO2 emissions, despite their low per capita emission rate. However, this trend changed after 2010, with upper-middle income countries surpassing high-income countries to become the largest contributor to global CO2 emissions in terms of total emissions. The reason for this shift is not hard to find: some countries changed their income category and brought their high emissions to the new category as well. This is a crucial point to be considered, as without paying attention to the emissions from large population countries, CO2 emission will be uncontrollable with the period of economic development.

Emission by Fuel

How different CO2 emissions from U.S. and China by fuel type?
Figure 5. An innovative view of CO2 emissions comparison of five sample countries by fuel. All of them use either coal or oil as a major fuel. Brazil, Canada, and Russia use oil as the major fuel, while China and U.S. use coal as the major fuel. These differences in energy source reliance reflect each country's unique energy mix and the historical development of its energy infrastructure.

Figure 1 shows positive relationship between energy use and CO2 emissions. Here, we conduct Figure 5 to exam the detailed distribution of CO2 emissions by fuel type for five sample countries adopted different forest strategy (Figure 3). When comparing the CO2 emissions from different energy sources for each country, we can observe the following: For coal, China has the highest reliance on coal (75%), followed by the United States (41.6%), Russia (38.8%), Canada (28.5%), and Brazil (14.9%). This indicates that China's CO2 emissions are heavily influenced by coal consumption, which has significant environmental implications. For oil, Brazil has the highest reliance on oil (68.7%), followed by Canada (44.1%), the United States (38.2%), Russia (28.3%), and China (13.8%). Brazil's transportation and industrial sectors contribute significantly to its CO2 emissions due to their dependence on oil. Russia has the highest reliance on natural gas (29.8%), followed by Canada (24.3%), the United States (18.6%), Brazil (7.97%), and China (2.87%). While natural gas is a cleaner-burning fossil fuel, it still contributes to CO2 emissions, with Russia being the most dependent on gas among these countries.

In summary, China has the highest dependence on coal, Brazil has the highest dependence on oil, and Russia has the highest dependence on natural gas among the five countries. Brazil also has the highest percentage of CO2 emissions from non-fossil fuel energy sources and some industrial processes. These differences in energy source reliance reflect each country's unique energy mix and the historical development of its energy infrastructure.

This plot is an innovative view. The plot is generated using R Plotly package. The color matches the energy use in Figure 1. We have ordered the fuel types same across the countries. We used clip charts in the middle of each pie chart to indicate the major energy source of each country.

Conclusion

After conducting our analysis, we have come to the conclusion that a direct, simple relationship between emissions and forest cover cannot be visually confirmed. While increasing forest coverage theoretically helps to absorb CO2, there are other major factors at play that impact CO2 emissions, making the effect of forest coverage changes negligible. One such factor is the high usage of fuel, particularly coal, oil, and gas, which are the dominant fuel types burned by the sample countries. Interestingly, this factor does not appear to have any visible impact on the overall trend of global emissions during the selected time period.
However, when examining individual countries, such as China, it is clear that energy usage has shown a steady increase from 1990 to 2015, while others, such as the U.S., have remained at the same level. The high usage of fuel may also be linked to income levels, as upper-middle and high-income countries, whose major emissions are contributed by Electricity and Heat, have been responsible for most of the global emissions in the past three decades. In fact, higher total emissions have been positively related to emissions per capita. In summary, while forest coverage is important in the fight against climate change, it appears that other factors, particularly energy usage and income levels, have a more significant impact on global emissions trends.

Data visulization

Find our Code here